ISSUE-118: Allow representing samples of samples
samples of samples
Allow representing samples of samples
- State:
- CLOSED
- Product:
- Semantic Sensor Network Ontology
- Raised by:
- Raúl García Castro
- Opened on:
- 2016-12-20
- Description:
- There is already an issue (ISSUE-92) related to whether samples should go into SOSA or SSN. This issue is about allowing the representation of samples of samples (independently of the module where they are included).
The proposal is to declare Sample as a subclass of FeatureOfInterest.
The conversation leading to this, extracted from the emails is the following:
Raúl: "From the documentation a Sample is a FeatureOfInterest (shouldn't Sample be a subclass of FeatureOfInterest?). I also think that there is no need for a Sample class; I would just state that a FeatureOfInterest can have as a sample another FeatureOfInterest. In any case, unless some of these changes are made, the current model "does not allow" taking samples of samples."
Krzysztof: "This is a longer discussion. We removed subclasses from SOSA some time ago and there are many good reasons for the sample class (see our argumentation for a FeatureOfInterest class). I agree with the sub-sampling issue and the subclass as a solution. Let me delay this for now, nonetheless. We will get back to this. Again, well spotted."
Simon: "The documentation may need to be improved, but there is a strong need for a Sample class. Many (most?) practical observations are on a sample of the ultimate feature of interest. This may be a physical sample taken for ex-situ analysis ('Specimen'), a spatial subset (e.g. cross-section, transect) used to characterize something much larger, or a subset of a population. In every case the intention is that the sample is representative of something larger, which is the real thing of interest. But the observation process uses the sample as a proxy for the bigger thing. Now in some cases we can elide or ignore the distinction, but in practice it is often the case in real world practice that having a clear model which includes both the proximate- and ultimate- feature of interest of an observation, and the relationships between them made clear, is essential to disentangle the semantics. This was one of the key achievements of O&M, and should be preserved in the work here. "
Krzysztof: "Spot on, thanks Simon. I think Raul's point was to declare Sample to be a subclass of FeatureOfInterest to allow for samples of samples and I agree that this is indeed a good idea and a change we should make to SOSA."
Raúl: "Yes, having Sample as subclass of FeatureOfInterest solves the issue. Clearly, any FeatureOfInterest that isSampleOf another FeatureOfInterest will be a Sample, so giving a name to such class is good."
- Related Actions Items:
- No related actions
- Related emails:
- Re: issue-117: Comments on the SOSA and SSN implementations (from rgarcia@fi.upm.es on 2016-12-22)
- RE: issue-117: Comments on the SOSA and SSN implementations (from Simon.Cox@csiro.au on 2016-12-22)
- RE: issue-117: Comments on the SOSA and SSN implementations (from kerry.taylor@anu.edu.au on 2016-12-22)
- Re: issue-117: Comments on the SOSA and SSN implementations (from rgarcia@fi.upm.es on 2016-12-20)
Related notes:
No additional notes.
Display change log
